It might be expected that in a disease like epilepsy, which shows itself in the definite if irregular recurrence of convulsive phenomena, the influence of diet might have been accurately ascertained. Yet it is apparently of so little consequence, in the opinion of those who have studied it and investigated it most closely, that in Sir William Gowers' classical work, which investigates over 3,000 cases, the subject of dietetic treatment is discussed and dismissed in little over a page. And when we consider our ignorance of the causes of epilepsy, this is scarcely to be wondered at. We know the disease only in its manifestations, the occurrence of convulsive seizures or attacks of unconsciousness of varying duration, and we know little or nothing of its underlying causes. Apparently, the necessary condition of the epileptic fit is an unstable state of certain cells in the brain cortex, a tendency in those cells to discharge, so as to give rise to convulsive movements of the muscles which they subserve, or to discharge in such a way as to leave other cells on a lower level of development free from the restraining influence of the higher cells. Further, these irritable cells seem to act as a fulminant, and so cause discharge in cells which are contiguous to them, cells in every respect healthy, yet which are induced to discharge by their unstable and excitable neighbours. It can readily be understood that their excitability is rendered obvious by different conditions.

There seems to be little doubt among different observers that emotional disturbance is one of the most common excitants of an epileptic fit. Any fright, any unusual excitement, any depressing emotion may determine a fit; yet it would not be correct to say that such an emotion might cause it. Similarly, a blow or a fall, that is, a physical shock as well as a psychical effect, may determine an attack, and so may any influence which modifies the circulation either by stimulating or depressing it. And this consideration probably explains the occurrence and recurrence of so-called nocturnal fits of epilepsy. Nothing is more striking, in investigating such nocturnal attacks, than the fact that, although they nearly always occur during the night, there are occasional diurnal attacks. Yet, on investigation, these diurnal attacks will be found to occur almost invariably during sleep, a chance sleep in which the patient has indulged during the day. So that such attacks of what is usually called nocturnal epilepsy are much more correctly to be described as attacks of epilepsy occurring during sleep, and the fact that they do so is, in all likelihood, to be ascribed to some modification of the circulation which occurs during sleep, giving rise to a condition of things which leads to a discharge from the unstable cortical cells. A curious confirmation of this view of epileptic fits occurring during sleep is to be found in the marked efficacy of cardiac stimulants like digitalis and nux vomica, in combination with bromides, in the treatment of those attacks.

Toxic conditions also are frequently an effective agent in evoking epileptic fits. Absinthe we know in experimental researches as one of the most effective producers of convulsion. It is not unlikely that, taken, as it frequently is, as a pick-me-up or appetizer, it may have a similar tendency to cause discharge in cells already predisposed. So also in uraemia, convulsions occur in no way distinguishable from epileptic convulsions. These two - the poison of absinthe and the poison of uraemia - are characteristic examples of exogenous and endogenous poisons producing convulsions. One cannot help asking the question - Does diet, or does any article of diet, especially in excess, produce some poison which may conceivably act in the same way as either of those? Much has been said and written about uric acid and its effects in producing nervous symptoms. Latterly, the pendulum has swung in the opposite direction, and now carbo-hydrates and purin bodies are the fashionable noxious agents. At different times, and by different observers, milk diet, a diet of milk, starchy food and vegetables, a purely vegetarian diet, a purin-free diet, and an ordinary mixed diet of bread and milk and vegetables have been vaunted as the diet most suitable for epileptics. The only point upon which all observers seem to be agreed is, that an excessive meat diet is bad for epileptics, just as it is probably, injurious for other people.

There are certain general considerations to be dealt with in discussing the diet suitable to any particular disease such as epilepsy. First of all must be considered the individual idiosyncrasy. There is no doubt that some patients can take with impunity more animal food than others. The same is true, in a less degree, of starchy food. Some patients, even epileptics, are better with a considerable quantity of animal food : others are undoubtedly best without it and thrive on milk, eggs and vegetables. The vegetarian diet alone is, in my experience, bad. It leads to weakness, anaemia, and general debility, and through this, not infrequently, to an increase in the number of fits. But in reference to any dietary it must always be remembered that an epileptic is very often not only hungry and greedy, but frequently voracious, and that much care and judgment have to be used in reference to his food, no matter what diet is decided upon. My own experience has led me very definitely to believe that epileptics, as a rule, are best with a good mixed diet of meat, milk and vegetable food. I am also strongly of opinion that care must be used to prevent the patient from overe-ating, and to ensure that the food shall be good, simple, and well cooked. Beef, as a general rule, is to be avoided, but mutton, chicken, game, milk, and starchy food may be given in fair quantity. Pastry should be avoided. Cheese may be given in moderate quantity, but alcohol is practically inadmissible. A purin-free diet has lately been advocated, but is by no means a panacea, though probably useful in certain cases, and in those probably only for a certain time. The exclusion of salt has also been urged, especially in patients taking bromides. It has been said that such an exclusion tends to make bromides more effective in their action, but this is doubtful, and in the series of cases which I observed at Queen Square some years ago it could not be stated that patients deprived of salt did better with similar doses of bromide than did those who took salt regularly. Yet in reference to this also, considering the claims that have been made for the effects of this omission of salt, it is probable that individuals do occasionally progress better under such a regime.

In some cases of epilepsy the fits occur only in the early morning, often while the patient is dressing. When this is the case, the patient should have a light meal - a cup of milk and a biscuit, or a cup of tea and a biscuit, or some coffee and milk - half an hour before getting out of bed. This meal should be immediately followed by a dose of bromide; and in such cases, which are by no means uncommon, this mode of treatment is usually crowned with complete and permanent success.

My experience, then, of diet in epilepsy has led me definitely to conclude that it is necessary, in order to get the best results, to treat each case of epilepsy on its merits. In certain cases it will be found best to give a light diet, consisting of milk, eggs, starchy food, and vegetables, and whatever dose of bromide is necessary to control the attacks. In other cases - and these I believe to be so numerous as to constitute a large majority - a mixed diet, of meat, fish, milk, eggs, and vegetables, well but simply cooked, will be found best: while in others it may be found advantageous to give a dietary as free as possible from purin bodies. It is impossible to lay down a definite rule for all epileptics, yet in the majority, a simple mixed diet, by no means too abundant, will be found best. In some cases the omission of common salt may be found to allow the bromide which is given to act with advantage; but in this, even more than in most morbid conditions, one has to remember that the patient has to be treated, as well as the disease.